JOURNAL OF OPTOELECTRONICS AND ADVANCED MATERIALS Vol. 12, No. 8, August 2010, p. 1752 - 1758
Chromium oxides mixtures in PLD films investigated by
Raman spectroscopy
O. MONNEREAU, L. TORTET, C. E. A. GRIGORESCUa*, D. SAVASTRUa, C. R. IORDANESCUa, F. GUINNETON,
R. NOTONIERb, A. TONETTOb, T. ZHANGc, I.N. MIHAILESCUd, D. STANOId, H.J. TRODAHLe
Laboratoire Chimie Provence and 3SCM, Universite de Provence, Marseille, France
a
National Institute R&D Optoelectronics INOE 2000, P.O. Box MG-5, Str Atomistilor No 409, Magurele, Judet Ilfov,
77125, Romania
b
SCM, Universite de Provence, Marseille, France
c
Renishaw plc, Old Town, Wotton-under-Edge, Gloucestershire GL12 7DW United Kingdom
d
NILPRP, Str Atomistilor No 409, Magurele, Judet Ilfov, 77125 Romania
e
Victoria University, PO Box 600, Wellington, New Zealand
The composition of Pulsed Laser Deposited thin films from Cr8O21 targets has been investigated by Raman spectroscopy.
To this purpose we have prepared the chromium oxides Cr8O21, Cr2O5, CrO2 and Cr2O3 in bulk form to provide standard
Raman spectra for characterisation of the samples. The range of oxides has been prepared by thermal decomposition of
CrO3 following the thermo-gravimetric analysis. CrO2 was obtained by two other specific ways, i.e. hot pressing of Cr8O21
and hydrothermal synthesis. Structural characterization was performed by X-ray diffraction, SEM, IR and Raman
spectroscopy. Raman investigation is demonstrated to have the capacity to identify CrO2 as well as trace concentrations of
secondary phases in the PLD thin films. The Raman measurements have led us also to the conclusion that PLD of Cr8O21
targets yields in thin films containing either mixtures of (Cr2O5, CrO2 and Cr2O3) or, under particular conditions of oxygen
pressure and substrate temperature, of (CrO2 and Cr2O3) only.
(Received July 14, 2010; accepted August 12, 2010)
Keywords: Chromium oxides, PLD, Thin films, Raman spectroscopy, Chemical synthesis
1. Introduction
As a new trend for a better use of the electron (both
the charge and the spin), “spintronics” is evolving along
with specific material systems such as the half metals
(HM) [1, 2]. The most popular HM systems to date can be
divided into two major groups: a) the half- and fullHeusler alloys [2] and b) various ferromagnetic oxides [3],
with CrO2 predicted to be one of the most promising,
showing fully spin polarized conductivity even at ambient
temperature [4]. However, chromium forms many
competing sub-oxides from which chromium dioxide is
metastable [5] at atmospheric pressure, decomposing into
Cr2O3 when heated [6,7]. Therefore, it is necessary to find
reliable characterisation protocols for these films, capable
of establishing the presence of even trace densities of
secondary Cr oxide phases. In this paper we examine the
capacity of Raman spectroscopy to provide such an
analysis and specifically apply it to an investigation of
pulsed laser deposition of Cr oxide films.
Although thin films of CrO2 have been successfully
prepared by chemical vapour deposition (CVD) [8] or
molecular beam epitaxy (MBE) [9-12] the route to film
growth by pulsed laser deposition (PLD) is not
straightforward, and requires that the starting material is
either CrO3 or Cr8O21 [12]. We thus review experimental
work that we have carried out on the range of chromium
oxides CrO3 to Cr2O3 with the aim to investigate PLD as a
potential technique for preparing thin films with a high
content of CrO2. In this investigation we have made
several attempts [13, 14] with Cr8O21 targets, which led to
PLD films consisting of mixtures of CrO2 and other
chromium oxides, such as Cr8O21, Cr2O5 and Cr2O3.
Characterization of both the films and the ablated targets
have aimed to identify the components of those mixtures
and consequently understand the processes that develop in
the CrxOy systems through the pulsed laser action [13]. Xray diffraction, SEM, IR and Raman spectroscopy have
contributed to this effort. XRD can give information about
fully crystalline phases with concentrations of over 1%wt.,
but when samples consist of mixtures with lowconcentration or amorphous secondary phases the
sensitivity of this technique is inadequate. Raman
spectroscopy offers an attractive alternative, which
furthermore has the potential to identify characteristic
bonding structures even in amorphous material [15]. It can
be further enhanced with the use of a range of excitation
wavelengths, which alters the balance between majority
and secondary spectra.
There are many reports in the literature of Raman [1621] and IR spectra [21-24], of the various Cr oxides and
several reports also of its use in identifying various phases
in composite Cr oxide material [25-27]. However, there is
insufficient data available on well-characterised single-
Chromium oxides mixtures in PLD films investigated by Raman spectroscopy
phase materials to understand all of the Raman signals we
have observed in our PLD films. With the exceptions of
CrO2 and Cr2O3 the spectra are still incompletely
characterized, in part because the compositions other than
Cr2O3 are unstable against loss of oxygen [28] and thus
difficult to prepare as pure single phase. It has thus been
necessary as part of this study to perform measurements
on large crystallites within samples of several Cr oxides,
and much of this paper is concerned with exactly that
background study.
2. Experimental
As outlined above this work concerns both
preparation of various materials and their Raman/IR
investigation. This section is arranged to describe the
preparation of Cr oxides by careful thermal decomposition
of CrO3 (section 2.1) and of CrO2 by two specific ways
(section 2.2). Section 2.3 describes the PLD targets and
thin-film growth. The Raman and IR spectroscopy is
covered in Section 2.4.
1753
2.1 Preparation of CrxOy (x=2,8;y=3,5,21) as pure
phases
We have prepared polycrystalline samples of the
different CrxOy phases starting from CrO3 (a commercial
powder of CrO3 -ACROS ORGANICS, Chromium (VI)
oxide 99.9%, ref: 214100010)). These phases were
obtained on the basis of the thermo-gravimetric analysis
(TGA) of the controlled thermal decomposition of CrO3 in
air (Fig. 1a). Powder of CrO3 is placed in an alumina
crucible set in a furnace and heated in air up to about
180°C. A low heating rate (1°C/h) is employed to achieve
the melting of CrO3 (Tm = 196°C). Then, a second slow
variation is performed around the temperature of
formation of the chromium oxide phase we are looking for
(260°C for Cr8O21, 330°C for Cr2O5 and 420°C for Cr2O3)
(Fig. 1b). The relative mass losses Δm/m starting from
CrO3 correspond to the expected values for Cr8O21, Cr2O3,
and Cr2O5.
We have obtained sheet-like crystals of about 20µm
width.
a
b
Fig. 1. a) thermo-gravimetric analysis of the controlled thermal decomposition of CrO3 in air b) thermal regimes
used to obtain Cr2O3 (a), Cr2O5(b) and Cr8O21 (c) from melted CrO3.
2.2 Synthesis of CrO2
Synthesis of CrO2 occurred by two specific ways: i)
hydrothermal and ii) hot-press;
i) the hydrothermal synthesis of chromium dioxide
powders is widely described in literature [29]. The CrO3
precursors in contact with water (providing both pressure
and a reaction environment) produce a partial dissolution
of the material forming chromium acid under ambient
conditions. This mixture was sealed in a gold container
placed in an autoclave under 500 bar external water
pressure. The process leading to CrO2 developed between
430°C and 480°C. We obtained black whisker-like grains,
with strong metallic reflections. Small but varying
amounts of Cr2O3 were found in the final product.
ii) a CrO2 target has been obtained by annealing
powder of Cr8O21 under a pressure of 0.074GPa. The
process developed 8h at 245°C. The pellet contained traces
of Cr2O5.
2.3 PLD targets and corresponding thin films
The detailed preparation of the Cr8O21 PLD targets as
well as of the corresponding PLD films are reported
elsewhere [13, 14, 30]. Briefly, pellets (about 1.5 mm
thick) are prepared in a 13 mm diameter steel die, under a
0.3 GPa pressure during 15 minutes and then annealed 16
hours at T = 245°C. The targets were ablated with a pulsed
laser to deposit thin films onto various substrates, such as
single-crystalline sapphire and quartz, held at temperatures
between 300K and 600K in the different experiments. The
PLD runs were performed with an UV KrF*excimer laser
(τFWHM≥20ns, λ=248nm), with fluences of 1-2 J/cm2 and
under a dynamical pressure of oxygen within the range 120 Pa. A total of 64 PLD films have been produced and
investigated.
2.4 Raman and infrared measurements
The Raman spectra of crystalline CrxOy powders as
well as of the Cr8O21 PLD targets and corresponding films
1754
O. Monnereau, L. Tortet, C. E. A. Grigorescu, D. Savastru, C. R. Iordanescu, F. Guinneton, R. Notonier, A. Tonetto…
were taken with a Renishaw RM2000 micro-Raman
instrument, in the back-scattering geometry. All
measurements have been carried out at room temperature,
with a 633 nm laser source, delivering 0.5mW on the
sample’s surface. Our work showed that a higher incident
laser power modifies the surface of the films and possibly
the composition too. The calibration was monitored using
the silicon signal and peak positions were determined with
the conventional Lorentz function provided in GRAMS/32
(Galactic Industries) software. The spectral resolution at
the spectrograph is better than 2 cm-1 with 50x objective.
IR transmission of the bulk samples was measured using a
Brücker Equinox 55 spectrometer. The samples were
pressed within KBr pellets (99%wt of KBr). The
crystalline chromium oxides show IR activity in the range
350 -1100 cm-1.
3. Results and discussion
Although the purity of Cr2O3 samples was not in
doubt and their characterization went smoothly, the
tendency of Cr8O21 [30] - to come out mixed with Cr2O5
[16, 22] raised a few difficulties in our attempts to produce
these materials as pure phases. Nonetheless the XRD
patterns (Fig. 2) taken on powders of our crystalline
Cr8O21 (47-1312), Cr2O5 (36-1329) and Cr2O3 (38- 1479)
samples are in a very good agreement with the standard
diffractogrammes. However, XRD on CrO2 (43 - 1040)
hot pressed pellet show clearly the presence of Cr2O5. The
IR spectra of our chromium oxides samples shown in the
Fig. 3 are in a good agreement with already published data
[22-25]. Obviously, if traces of Cr2O3 had occurred in the
samples, they would have shown up in the IR spectra. It is
significant that the Cr2O5 from this work shows no trace of
Cr8O21 in comparison with previously reported data [16].
This may be due mainly to the fact that our samples
are large enough platelet-like crystallites (see SEM images
in Fig. 4) while the ones reported in ref. [16] were ceramic
pellets where phases intergrowth had most likely occurred.
Fig. 2. XRD patterns of powder samples from crystalline:
a) Cr8O21, b) Cr2O5, c) CrO2 and d) Cr2O3. The symbol *
in the spectrum c) signifies Cr2O5.
Fig. 3. The IR spectra of the chromium oxides samples:
a) Cr2O3, b) CrO2, c) Cr2O5, d) Cr8O21, e) CrO3
Fig.4. SEM images of platelet-like crystallites a) Cr8O21, b) Cr2O5, c) CrO2, d) Cr2O3
Chromium oxides mixtures in PLD films investigated by Raman spectroscopy
CrO2 has been predicted to have 4 IR active modes
[33], and Iliev [17] cites unpublished results of optical
conductivity measurements in the IR range showing 3
peaks. However, IR transmission measurements on our
CrO2 samples show no evidence of those modes, being in
agreement with earlier publications [24]. A possible
explanation of this result may be that the high absorbance
of CrO2 dominates the spectrum, masking features from
the other chromium oxides. Therefore, it is impossible to
see contributions from secondary phases in the IR spectra
of this compound. In all of these materials the secondary
phase content is small, but it is essential that its influence
on the Raman or IR spectra be monitored. It is particularly
important to monitor the presence of any features that
might result from a small density of Cr2O3 crystallites or
intergrowths, for this is the most stable of all Cr oxides,
and it is known to form in any case when these materials
are exposed to air for an extended period [33].
The Raman spectra of the crystalline Cr8O21, Cr2O5
and Cr2O3 are displayed in Fig.5 (a, b, and c) and the main
peak frequencies resulted from curve fitting are listed in
the table I. Their strongest features are consistent with data
published earlier [16,19, 27]. In view of the potential for
Cr2O3 intergrowths in any of these materials we choose to
discuss first the data from that sample. These are
1755
particularly clear, with the phase identified by strong
doublets at 550/610 cm-1 and 305/348 cm-1. In addition
these show weaker satellites at 524 and 397 cm-1 and a
weak line at 826 cm-1 .
Fig. 5. Raman spectra of a) Cr2O3, b) Cr2O5, c) Cr8O21 single
phase samples.
Table 1. Main Raman frequencies resulted from curve fitting.
Cr2O3
[cm-1]
Raman
freq.
305, 348, 397, 524, 550,
610, 826
Cr2O5 Raman freq.
[cm-1]
188, 214, 253, 269, 312,
327, 353, 387, 421, 446,
477, 629, 741, 838, 865,
880, 996
Neither the 397 [20], nor the 826 cm-1 lines seem to be
issued from the Cr2O3 structure. Whereas the 397 cm-1
feature is presumed to be responsible for the one-magnon
scattering [20], 826 has yet to be assigned [27]). However,
it is the pair of strong doublets that will signal the presence
of Cr2O3 in any multiphase film. Turning next to the
Raman features exhibited by Cr8O21 and Cr2O5 there are
strongly dominant lines at 899 cm-1 and 880 cm-1
respectively, corresponding satisfactorily to the 904 cm-1
and 884 cm-1 in the work [16]. There are in addition
weaker lines spread across the spectra, such as those in the
400-500 cm-1 range, that again correspond with earlier
data in reference [16], and in agreement with that report
these are seen at nearly two orders of magnitude lower
level than the dominant lines near 900 cm-1. CrIII oxides
(CrIII-O6) exhibit peaks in the range 303-609 cm-1 [19],
which again agree with the values of the Raman and IR
wavenumbers (see Fig. 3) we have obtained for our
samples Thus it is clear that these phases should be
associated with strong lines at 900 cm-1 (Cr8O21) and 880
cm-1 (Cr2O5) at least under 633 nm excitation.
All Raman spectra of CrO2 given in literature were
taken on epitaxial films [18] to our knowledge. The main
Raman modes of CrO2 (458 and 570 cm-1) [17] are present
Cr8O21 Raman freq.
[cm-1]
246, 308, 344, 368, 397,
487, 549, 732, 878, 899,
1001
in the spectra of the two CrO2 samples (hot-pressed pellet
and whiskers from hydrothermal synthesis), which are
compared in the Fig.6.
Fig. 6. Raman spectra of CrO2. The main Raman modes
of CrO2 (458 and 570 cm-1) are present in the spectra of:
a) hot-pressed pellet and b) whiskers from hydrothermal
synthesis. The symbol * corresponds to Cr2O3.
1756
O. Monnereau, L. Tortet, C. E. A. Grigorescu, D. Savastru, C. R. Iordanescu, F. Guinneton, R. Notonier, A. Tonetto…
However, the Raman spectrum of the hydrothermally
produced powders shows some additional peaks, which
correspond to Cr2O3. The explanation is straightforward
through the preparation process (see section 2.2.) on one
hand, and through the loss of oxygen from the powders in
contact with atmospheric air, on the other hand.
This second reason can also explain the observation of
Cr2O3 peaks in the Raman spectrum of the hot-pressed
pellet. Raman scattering is a most surface sensitive
technique in comparison with XRD and IR measurements.
Before attempting to interpret the Raman spectra
taken on PLD films it is important to discuss the influence
that a disordered or nanocrystalline structure might have
on the spectra. Firstly small crystals of diameter d permit
the Raman activity of non-zone-centre modes, out to wave
vectors of about π/d. That will lead to both broadening of
the lines and either up or down shifts, depending on the
sense of the phonon dispersion [28]. In the absence of any
vibration mode calculations it is impossible to predict in
which sense the shift will occur, but it can be expected that
the magnitudes of broadening and shifts will be similar. In
addition there can be expected to be defect lines
originating from intergrain regions or even within strongly
defected crystals. One obvious source is the CrVI-O
terminating structures that occur also in some of the
ordered oxides and lead to the strong lines in the 800-1000
cm-1 region [16].
Armed with CrxOy spectra as described above we have
attempted interpretations of the Raman measurements of
PLD films (Fig. 7) deposited from Cr8O21 targets onto
sapphire substrates held at Ts 300 (a), 450 (b) and
600K(c).
cm-1 is not identifiable as originating from any of the
phases we have prepared in bulk. We speculate that it is
associated with terminating CrVI-O structures mentioned
by Maslar [27], in intergrain regions. It is interesting to
note that it appears also as a minor peak in our spectra
from bulk Cr2O3 samples, but the feature in these PLD
films is much too strong in relation to the pair of doublets
to assign it to that phase. Thus, the two films grown at the
corresponding substrate temperatures of 300K and 450K
are mixtures of Cr2O5 and Cr2O3.
The 600 K film consists of a mixture of CrO2 and
Cr2O3, with CrO2 leading to the prominence of the B2g
mode at 682 cm-1 and the large broad feature lying across
580-150 cm-1[17, 28]. The CrO2 in the film is in high
concentration but clearly nanocrystalline [28], as signaled
by the fact that these lines are prominent but much broader
than in single crystals. The second compound in the film is
Cr2O3, probably at a small concentration but well
crystallised once it is only this one coming out in the XRD
spectrum [13]. Magnetic measurements confirmed CrO2 in
this film.
4. Conclusions
The nearly single phase Cr8O21, Cr2O5, CrO2 and
Cr2O3 have been prepared in bulk form and used as
standard samples to establish the Raman spectra to be
expected from various phases within CrxOy PLD films.
Clear Raman signatures of the various oxides have been
identified and used to investigate the phases present in
PLD films. From the point of view of PLD, the laser
interaction with the Cr8O21 targets seems to produce the
lower chromium oxides (see Fig. 1a), which are then
deposited or even formed on the substrate under the
specific conditions of the PLD run, such as oxygen
pressure, substrate temperature, fluence. Raman
spectroscopy and its sensitivity to some minority phases
have been demonstrated as particularly useful as an aid to
characterisation of PLD films. However, a particular
difficulty can arise when the material being sought in a
mixture is a reasonable conductor (CrO2) and the other
phases are insulators (Cr2O3, for example), as it is the case
in our PLD films.
Acknowledgements
Fig. 7. Raman spectra of PLD thin films deposited at
Ts= 300K (a), 450K, (b) and 600K, (c)
The * correspond to Cr2O3.
It is seen immediately that all three spectra have the
pair of doublets associated with Cr2O3, so the films all
contain a low concentration of that phase. It is notable that
none of the PLD spectra shows any weight at 900 cm-1
suggesting that there is no measurable level of Cr8O21 in
the films. It is likely that the line at 864 cm-1, which does
indeed carry substantial weight at 880 cm-1, signals the
presence of nanocrystalline Cr2O5, but its partner at 826
Part of this work has been supported from the EC
Contract G5-RD-CT-2001-00535 –FENIKS. Later support
is acknowledged from MacDiarmid Institute of Advanced
Materials and Nanotechnology SCPS Victoria University
of Wellington, New Zealand (CEAG, 2007, 2008) and
PNCDI 2, Project FUNFOTON (2008) Romania
References
[1] C. Hordequin, J. P. Nozieres, J. Pierre, J. Magn.
Magn. Materials 183, 225, (1998).
[2] L. H. Leuken, R. A. De Groot, Phys Rev
B 51,7176, (1995).
Chromium oxides mixtures in PLD films investigated by Raman spectroscopy
[3] R.J. Soulen, J. M. Byers, M.S. Osofsky et al, Science
282, 85 (1998).
[4] P. Schlottmann, J. Appl. Phys. 95, 7471, (2004).
[5] B. L. Chamberland, CRC Crit. Rev. Solid State
Mater. Sci.7, 1,(1977).
[6] K.A. Wilhelmi, Acta Chem. Scand 22, 2565 (1968).
[7] G. Loithoir, Thesis., Paris, Orsay, (1965).
[8] K. A. Yates, W. R. Branford, F. Magnus, Y. Miyoshi,
B. Morris, L. F. Cohen, P. M. Sousa, O. Conde,
A. J. Silvestre, Appl. Phys. Letts. 91, 172504 (2007).
[9] P.G. Ivanov, S.M. Watts, D.M. Lind, J. Appl. Phys.
89, 1035, (2001).
[10] M. Rabe et al., J. Magn. Magn. Materials
211, 314, (2000).
[11] A. F. Mota, A. J. Silvestre, P.M. Sousa, O. Conde,
M. A. Rosa, M. Godinho, Int. Mat. Symposium,
Lisbon, Portugal, (March 2005).
[12] P. G. Ivanov, Ph. D. Thesis, Florida State Univ.,
(2002).
[13] F. Guinneton, O. Monnereau, L. Argeme, D. Stanoi,
G. Socol, I. N. Mihailescu, T. Zhang, C. Grigorescu,
H. J. Trodahl, L. Tortet, Appl. Surface Science
247, 139, (2005).
[14] D. Stanoi, G. Socol, C. Grigorescu, F. Guinneton,
O. Monnereau, L. tortet, T. Zhang, I. N. Mihailescu,
Mater. Sci. Eng. B 118, 74, (2005).
[15] Harish C. Barshilia, K.S. Rajam, Appl. Surface
Science 255, 2925 (2008).
[16] J. E. Maslar, W. S. Hurst, T. A. Vanderah, I. Levin, J.
Raman Spectroscopy 32, 201, (2001).
[17] N. M. Iliev, A. P. Litvinchuk H.G. Lee, C. W. Chu,
A. Barry, JMD Coey, Phys. Rev.B 60, 33,(1999).
[18] Ramakant Srivastava, L. L. Chase, Solid State
Com.11, 349, (1972).
[19] I. R. Beattie, TR. Gilson, J. Chem. Soc.
(A), 980 (1970).
[20] T. R. Hart, R. L. Aggarwal, B. Lax, Proc. Int. Conf.
light scattering in solids, (Paris - France),
174, (1971).
1757
[21] D. A. Brown, D. Cunningham, W. K. Glass,
Spectrochimica 24A, 965, (1968).
[22] Nasr E. Fouad, Bull. Fac. Sci., Assiut Univ.,
22(1-3), 55, (1993).
[23] T. A. Hewston and B. L. Chamberland, J. Magn.
Magn. Materials 43, 89 (1984).
[24] R. M. Chrenko & D.S. Rodbell, Physics Letters,
24A(4), 211, (1967).
[25] T. Yu, Z. X. Shen, J. He,W. X. Sun, S. H. Tang,
J. Y. Lin, JAP 93(7), 3951, (2003).
[26] J. Mougin, N. Rosman, G. Lucazeau , A.Galerie,
J. Ram. Spectr. 32(9), 739 (2001).
[27] J. E. Maslar, W. S. Hurst, W. J. Bowers,
J. H. Hendricks, M. I. Aquino, I.Levin, Appl. Surface
Science 180, 102, (2001).
[28] Yanping Chen, Kui Ding, Ling Yang, Bo Xie, Fengqi
Song, Jianguo Wan, Guanghou Wang, Min Han,
Appl. Phys. Letts. 92, 173112 (2008).
[29] O. Schäf, H. Ghobarkar, P. Knauth: «Hydrothermal
Synthesis of Nano-Materials» in « Nanostructured
Materials: Selected Synthesis Methods, Properties
and Applications», P. Knauth and J. Schoonman
(eds.), Kluwer, Boston, 23, (2002).
[30] L.Tortet, F. Guinneton, O. Monnereau, D. Stanoi,
G. Socol, I. N. Mihailescu, T. Zhang, C. Grigorescu,
Crystal Research and Technology
40(12), 1124, (2005).
[31] P. Norby, A. Norlund Christensen, F. Fjellvag,
M. Nielsen, J. Solid State Chem.94, 281, (1991).
[31] S. P.Lewis, P.B.Allen, T.Sasaki, Phys Rev
B 55(16),10253, (1997).
[33] R. Cheng, B. Xu, C. N. Borca, A. Sokolov,
C. S. Yang, L. Yuan, S. H. Liou, B. Doudin,
P. A. Dowlen, Appl. Phys. Lett.
79(19), 3122, (2001).
______________________________
*
Corresponding author: [email protected]